1. Field of the Invention
The present invention relates to a control apparatus for suitably selecting a predetermined region in an observed image region, a control method, and a program for implementing the method.
2. Description of the Related Art
In recent years, with the spread of the internet and the move from analog to digital in products such as televisions and cameras, people have become capable of dealing normally with higher levels of functionality than in the past.
For instance, people are able to select predetermined regions included in an image while observing images on a predetermined image display device such as a computer terminal, a mobile telephone, or a digital TV, and thereby perform an operation that is preset for the selected region. Operations such as trimming to form trimmed regions and operations to cause display at a lower resolution can be performed in the same way.
In one technology for selecting predetermined regions, two points are freely indicated on the displayed image screen, and used as diagonally opposing corners for defining the range of a rectangular region (see, for instance, Japanese Laid-Open Patent Publication (Kokai) No. 2001-145094). In another technology, a central point of the region is freely indicated, and enlargement/reduction with the central point at the center is then specified (see, for instance, Japanese Laid-Open Patent Publication (Kokai) No. 2004-172655) In a further technology, rectangles of predetermined sizes are freely moved to select a region (see, for instance, Japanese Laid-Open Patent Publication (Kokai) No. 9-116886).
Further, for improving operability when specifying human subjects, there has been proposed a technology in which, the subjects' faces are detected using image recognition techniques, and an arbitrary region in a plurality of regions are selected automatically or by a user when the plurality of regions are detected (see, for instance, Japanese Laid-Open Patent Publication (Kokai) No. 2005-86682).
However, since the increased functionality of digital products has tended to increase the number of operations buttons on remote controllers and the like, unseasoned users' confusion when attempting to perform a desired operation has become a problem. An increase in operability is therefore required as well as the increase of functions.
For instance, in the technology disclosed in the above-described Japanese Laid-Open Patent Publication (Kokai) No. 2001-145094, it is assumed that the image display device is a computer terminal and the region is selected mainly using a mouse. Hence this technology is not suitable for selecting regions using operations buttons of an image display device such as a mobile telephone, or a digital TV.
In the technology disclosed in the above-described Japanese Laid-Open Patent Publication (Kokai) No. 2004-172655, two operations are necessary, the operation to indicate the central point and the operation to adjust the range of the area. This makes the overall operation complicated.
Further, in the technology of the above-described Japanese Laid-Open Patent Publication (Kokai) No. 9-116886, although a plurality of rectangles are prepared for selecting the region, when no rectangle of a size equivalent to the region to be selected is available, a region that is one size larger must be selected. In this case, areas not required by the original selection are included in the selected region.
Also, since the shape the rectangles for selecting the region are predetermined, it is not possible to specify an intricately shaped region, such as a region defined by the outline of a house or the like.
In the above-described Japanese Laid-Open Patent Publication (Kokai) No. 2005-86682, image recognition is performed, and it is therefore possible to extract a region such as the above-described outline of the house. In reality, the region to be selected may be a person rather than a house, and any size and shape is conceivable as the selected region.
In the above-described Japanese Laid-Open Patent Publication (Kokai) No. 2005-86682, the image extraction is performed with a limited of predefined targets for image recognition, and it is sometimes the case that the region to be selected cannot be flexibly switched.